base_model: tokyotech-llm/Llama-3.1-Swallow-8B-Instruct-v0.1
models:
  - model: tokyotech-llm/Llama-3.1-Swallow-8B-Instruct-v0.1
    # No parameters necessary for base model
  - model: NousResearch/Hermes-3-Llama-3.1-8B
    parameters:
      density: 0.525
      weight: 1
merge_method: dare_ties
tokenizer_source: union
dtype: bfloat16

💻 Usage

python

!pip install -qU transformers accelerate

from transformers import AutoTokenizer
import transformers
import torch

model = "AELLM/llama-3.1-swallow-hermes-8b-v0.1"
messages = [{"role": "user", "content": "あなたはタイムトラベラーのAIアシスタントだ。タイムパラドックスを引き起こすことなく、異なる歴史的時代に溶け込むにはどうすればいいのか？"}]

tokenizer = AutoTokenizer.from_pretrained(model)
prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
pipeline = transformers.pipeline(
    "text-generation",
    model=model,
    torch_dtype=torch.float16,
    device_map="auto",
)

outputs = pipeline(prompt, max_new_tokens=256, do_sample=True, temperature=0.7, top_k=50, top_p=0.95)
print(outputs[0]["generated_text"])